Analysis and pattern recognition of blast furnace burden surface based on multi-radar data

Iron making is the first stage and also an important part in steel making process, which will bring the problem of energy efficiency and economic benefits. In this process, the distribution of the blast burden impacts greatly on the production of BF (blast furnace). Therefore, the prediction of furnace burden distribution in furnace throat will play an important role in the control strategy of furnace burden. Based on the data from the multi-radar, the blast burden curve can be formed. Feature extraction and classification based on different curves (i.e. different pattern) can be made by using BP neural networks. The work proposed in this paper will be a guidance for the future research of burden surface based on the data of phased-array radar.